Project summary Background: The bacterium that causes tuberculosis (TB), Mycobacterium tuberculosis (M.tb), has infected one third of the world's population. The majority (90%) of infected persons never progress to disease. In absence of disease, M.tb can not be isolated and infection is presumed when a memory T cell response to M.tb antigens is detected in blood (QuantiFERON TB Gold In Tube (QFT), an in vitro IFN? release assays, IGRA) or a delayed hypersensitivity reaction develops upon injection of M.tb antigens in the skin (tuberculin skin test, TST). Memory T cells are thought to persist in absence of stimulation, but in chronic infections their frequency, phenotype and function are affected by antigen load. In some individuals TST and QFT responses become undetectable over time. In natural history studies in humans and guinea pigs spontaneous TST reversions were associated with lack of TB disease and clearance of M.tb infection, suggesting that immune responses associated with TST reversion may eradicate the infection. Clinical and immunological significance of QFT reversion are unclear, complicating the interpretation of serial testing. Objective: To define the immunological determinants of QFT reversion in high TB burden settings. Design: From a large prospective adolescent cohort study, we selected 4 groups of participants based on 4 serial QFT tests conducted 6 months apart: 1. Reverters (QFT +/+/-/-); 2. Converters (QFT -/-/+/+); 3. Persistent negatives (QFT -/-/-/-); 4. Persistent positives (QFT +/+/+/+). Aim 1: To define antigen specificity, functional and phenotypic characteristics of classical T cell responses associated with QFT reversion. QFT reversions may be due to technical variability of the assay. In alternative, we hypothesize that QFT reversion may be associated with recognition of different M.tb antigens and/or production of cytokines other than IFN?. We will confirm QFT reversion including alternative M.tb antigens to stimulate cells and using flow cytometry, a technique that enables detection of multiple cytokines as well expression of phenotypic markers at single cell level. Aim 2: To determine if innate and unconventional lymphocyte responses to M.tb change upon QFT reversion. Loss of persistence of memory T cells could be associated with transient infection controlled by innate immune cells. We will determine whether immune responses to mycobacteria mediated by unconventional and non-T cells are associated with QFT reversion by flow cytometry. Relevance: QFT is increasingly used in alternative to TST to define M.tb infection status and, recently, as end-point assay for vaccine efficacy trials of prevention of M.tb infection. Our results will help clarifying whether QFT reversions are likely due to technical variability, and in such case will be invaluable to design better diagnostics. In alternative, we will identify immune responses potentially associated with non- persisting M.tb infection, which could be immediately validated in ongoing efficacy trials.